ZoneMikel's Blog

Abstract

Even though I have studied this subject several times it still continues to annoy me.  This will be a in depth disscussion (kinda for myself) of the signal levels and such. Im wondering if CCS C is inverting the signal coming out of the pic so that its ready for consumption by the db9.

This straight out sort of conversion works off of the fact that most rs232 with short range cables can work just fine with +/- 3 volts. So if you use a transistor to invert the pic signal and a resistor to sink the current when the line is 0 you will get +/-3 volts to drive the rs232 line.  So basically we invert the data coming out of the pic and add a resistor to that so when the pic outputs a 1 it is inverted to a 0 then sunk to the resistor to make -3 ish. This is best looked at with a table

Table 1

Pic out | RS-232 In
0volt    | +3volt
5volt     | -3volt

This is because for the pic a ‘1′ is 5v and for the rs-232 a ‘1′ is less than -3volts, likewize for the pic a ‘0′ is 0v and for rs-232 it is above 3v. So its pretty easy (in theory) to convert from the pic to rs-232, converting from rs-232 to pic levels is more tricky though. But, enough theory and such lets see what we actually get through the scope.

Results

Here I will show you (and myself) the results i get from rs232 and pic while sending the same characters.  Hopefully after looking at everything and putting it in perspective we will be able to clearly see whats going on, or we will realize what we are doing wrong.

Setting the scope (soundcard scope)

I’m going to send ‘A’ then ‘a’ at 1200 baud via the pic, and via rs-232, we will be using eight bits with one stop bit. This makes 9 bits total so 1200/9 gives ~133 Hz for each bit. A less confusing way of figuring this out is: we have 1200 bits per second so thats 1/1200 8.3e-4 seconds or (*1000) .83 milliseconds for each bit. We know that we are getting 9 bits so to see the “whole thing” we use .83ms*9bits= 7.5ms.  When i set my scope to 8ms i see nothing so i will just use about 60ms then zoom in.

Getting the shot

If you turn on your speakers and your using the soundcard scope you will be able to “hear” the bits go through. I use this as a metronome and i hit the run/stop button to get a good shot of what was sent. Then in the mid bottom i go to the “measure” setting and put it on cursor time. Then i get the blue lines and drag them from the start till about 130hz. This is what i get

So where i have the blue lines is the first char ‘A’ and the rest is the second char ‘a’. From http://www.asciitable.com/ we know that ‘A’ is 65D or 1000001 in binary. Looking at the output we can see we have something like 10???01, we know those four bits so the rest must be 9-4=5. But, I think there is always a “start” bit so that would make it 10 bits we really have so 10-4=6, and that chunk in the middle is all 1’s so 1011111101 is what we have which looks like

start bit | (data)’ | stop bit

I know we cant tell with this one but the data is sent lsb first, and yes that is the inverse of the data. So, where there was a 1 there is a 0 and so forth. This is what we get out of the pic, there is no manipulation its just “putc(‘A’)” so either it should be inverted, or ccs c is inverting it for us. Looking at the ‘a’ we see it is 1100001, and we are getting 10???001. Since we can make out 5 bits, 5 are left that chunk is 1’s so we get 1011111001. Which is just like before start/stop bits and lsb first inverted data.  You dont have to figure out just by looking at it what the bits are you can start from the first bit and set f=134/9 and see what each bit is if need be.

Doing the same thing with RS232

It wasnt as easy to get the shot with rs232 but i opened hyperterminal set bits to 8 stop bits to one and baud to 1200. Then while hitting ‘A’ i hurried up and hit run/stop to get this.

From before we know that ‘A’ is 1000001, and yes all that below the red line is negative so we get “1011111101″, assuming that second bit is supposed to be above 3v (what else?). Also i already inverted everything from rs232 to ttl in my head. So we see that here too we get the data portion inverted.

Well, this is what we are getting, but according to sites like http://www.winpicprog.co.uk/pic_tutorial7.htm we should be getting something compleatly different, mainly inverted stuff. But im going to see terminator “salvation” in a hour so ill have to come back to this